Token导航 LogoToken导航TokenDH.com
研究检索敏感数据github未标认证来源可访问clear审计提醒

daily-report日报

Agent Skill

daily-report 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

2,470

周安装

105

GitHub Stars

9

下载量

865
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

3

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:daily-report(日报)
来源仓库:https://github.com/i9wa4/dotfiles
仓库路径:skills/daily-report
安装命令:
npx skills add https://github.com/i9wa4/dotfiles --skill daily-report
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/i9wa4/dotfiles --skill daily-report

简介

daily-report 汇总 GitHub 与 Jira 活动,以 Issue 形式发布日报。

  • 自动拉取最近 24 小时操作记录,支持自定义时间范围。
  • 需安装 gh CLI、jq 和 acli 工具链方可正常运行。
  • 涉及账号凭证与项目访问权限,建议使用最小必要授权原则。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

Daily Report Skill

Summarize @i9wa4's GitHub and Jira activities and post as a GitHub Issue.

1. Prerequisites

  • gh CLI installed
  • jq installed
  • acli installed (for Jira)

2. Workflow

2.1. Get GitHub Activities

Use dedicated script to fetch activities. Defaults to fetching 24 hours of activities from current time (calculated in UTC).

2.1.1. Script Location

${CLAUDE_CONFIG_DIR}/skills/daily-report/scripts/get-activities.sh

2.1.2. Command Examples

# Default: 24 hours ago to now (local repos only via ghq list)
${CLAUDE_CONFIG_DIR}/skills/daily-report/scripts/get-activities.sh --no-url

# Specify hours: N hours ago to now
${CLAUDE_CONFIG_DIR}/skills/daily-report/scripts/get-activities.sh --hours 48 --no-url

# Specify datetime directly (ISO8601, UTC)
${CLAUDE_CONFIG_DIR}/skills/daily-report/scripts/get-activities.sh --from 2025-12-16T15:00:00Z --to 2025-12-17T15:00:00Z --no-url

2.1.3. Options

OptionDescription
--no-urlOutput without URLs (prevents mention notifications)
--hours NFetch activities from N hours ago to now
--fromStart datetime (ISO8601, e.g., 2025-12-17T00:00:00Z)
--toEnd datetime (ISO8601, e.g., 2025-12-17T23:59:59Z)
--hostnameGitHub Enterprise Server hostname
--exclude-ownerExclude repos by owner (comma-separated, default: i9wa4)
--include-personalInclude personal repos (overrides --exclude-owner)

2.1.4. Output Format

### repository-owner/repository-name

- [Issue] Issue title
- [IssueComment] Issue title
- [PullRequest] PR title
- [PullRequestComment] PR title
- [ReviewedPR] PR title (PRs reviewed by me, excludes my own PRs)

Note: Use --no-url option to omit URLs and prevent mention notifications. Note: Personal repos (i9wa4/*) are excluded by default.

2.2. Get Jira Activities

Use acli to fetch Jira activities.

# Today's activities
acli jira workitem search \
    --jql "updated >= startOfDay() AND (assignee = currentUser() OR reporter = currentUser()) ORDER BY updated DESC" \
    --fields "key,summary,status"

# Specific date range
acli jira workitem search \
    --jql "updated >= 'YYYY-MM-DD' AND updated < 'YYYY-MM-DD' AND (assignee = currentUser() OR reporter = currentUser()) ORDER BY updated DESC" \
    --fields "key,summary,status"

2.3. Get Meetings from Google Calendar

Fetch today's meetings from Google Calendar via Slack DM.

2.3.1. Prerequisites

Environment variables required:

export SLACK_GCAL_DM_URL="https://app.slack.com/client/E.../D..."
export SLACK_MCP_XOXC_TOKEN="xoxc-..."
export SLACK_MCP_XOXD_TOKEN="xoxd-..."

2.3.2. Fetch Messages from Last 24 Hours

# Extract channel ID from URL
CHANNEL=$(echo "$SLACK_GCAL_DM_URL" | sed -n 's|.*/client/[^/]*/\([^/]*\).*|\1|p')

# Fetch messages
FILE=$(mkoutput --dir tmp --label output)
curl -s -X POST "https://slack.com/api/conversations.history" \
  -H "Authorization: Bearer $SLACK_MCP_XOXC_TOKEN" \
  -H "Cookie: d=$SLACK_MCP_XOXD_TOKEN" \
  -d "channel=${CHANNEL}" \
  -d "limit=200" > "$FILE"

# Filter messages from last 24 hours
NOW=$(date +%s)
YESTERDAY=$((NOW - 86400))
jq -r --argjson yesterday "$YESTERDAY" \
  '.messages[] | select(.ts | tonumber > $yesterday)' "$FILE"

2.3.3. Extract Today's Schedule

Look for the "Today" summary message in 24-hour messages. This message contains all meetings for the day in attachments array.

Key indicators:

  • Message text starts with *Today*-<!date^...
  • Attachments contain meeting info with Unix timestamps

2.3.4. Parse Meeting Times from Timestamps

Meetings use Unix timestamp format: <!date^UNIX_TIMESTAMP^{time}|...>

Convert timestamps to local time:

date -r UNIX_TIMESTAMP "+%H:%M"

2.3.5. Output Format

Extract from attachments:

  • Meeting title: From link text (e.g., |Meeting Name>*)
  • Start/End time: Convert Unix timestamps to HH:MM format
  • Skip cancelled meetings (indicated by strike-through in separate messages)

Format as: - Meeting name (HH:MM-HH:MM)

2.4. Create Draft

Create file:

FILE=$(mkoutput --dir tmp --label "daily-$(whoami)")

NOTE: Keep command blocks in "Today's AI Activities" section. Execute commands and paste results below the command block.

Template:

---
title: "YYYY-MM-DD $(whoami)"
labels:
  - name: "日報"
    color: "0887b9"
---

## 1. Reflection

## 2. Today's Activities

### 2.1. GitHub

Organize gh-furik output. Classify as follows:

#### 2.1.1. Created Issues

- [repo-name] Issue title

#### 2.1.2. Created PRs

- [repo-name] PR title
- [repo-name] PR title (merged)
  - Add supplementary comments indented

#### 2.1.3. Reviewed PRs

- [repo-name] PR title

#### 2.1.4. Commented Issues/PRs

- [repo-name] Issue/PR title

### 2.2. Jira

- [KEY-123] Issue summary (status)

### 2.3. Meetings

- Meeting name
  - Supplementary comments

2.5. Wait for User Edit

Display draft and wait for user edits. User adds reflections.

2.6. Post Issue

Post with gh issue create:

gh issue create --title "YYYY-MM-DD $(whoami)" --label "日報" --body "$(cat <<'EOF'
[body]
EOF
)"

Display Issue URL after posting.

3. Important Rules

  1. No direct links: Do not link PRs/Issues directly (triggers mentions), use titles only
  2. PR/Issue numbers: Use (# 123) with space, not (#123). Without space, GitHub auto-links to unintended repos.
  3. Draft review: Always show draft to user and wait for edits
  4. Issue posting: Post after user confirms draft is ready

4. Output Type Classification

Classify script output as follows:

Output TypeClassification
IssueCreated Issues
PullRequestCreated PRs
ReviewedPRReviewed PRs
IssueCommentCommented Issues/PRs
PullRequestCommentCommented Issues/PRs

Note: Consolidate multiple comments on same Issue/PR into one. Note: ReviewedPR only includes PRs authored by others (my own PRs excluded).

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Claude Code

31.72%
按下载量换算274

windsurf

22.8%
按下载量换算197

trae

18.82%
按下载量换算163

OpenCode

13.87%
按下载量换算120

Codex

7.95%
按下载量换算69

Antigravity

3.55%
按下载量换算31

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。

来源信息

继续浏览同类 Skills